See Enterprise GenAI Outlook 2025 Survey Results
[ Overview ]

About the Video

Welcome to another episode of AI Weekly Breakthroughs, where we explore key developments in the AI world.

Grok 2 Released

XAI, Elon Musk’s AI company, has released Grok 2 in beta. This new model shows performance improvements over Grok 1.5 in:

– Chat capabilities

– Coding proficiency

– Reasoning abilities

Additionally, XAI has introduced Grok 2 Mini, a smaller, more focused sibling of Grok 2, reflecting a trend seen in the industry. Companies are now releasing families of models: larger, highly capable models and smaller models designed for specific tasks at a lower cost.

Availability

Both models are available in beta on the X platform, and they will also be accessible through XAI’s enterprise API later this month.

Performance

On the LMSYS Benchmark, Grok 2 ranks fourth, outperforming:

– Cloud 3.5 Sonnet

– GPT-4 o Mini

However, it trails behind:

– GPT-4 o

– Gemini 1.5 Pro

Notably, the LMSYS Benchmark has a new leader. OpenAI has reclaimed the top spot with the latest August version of ChatGPT 4.08.

Unique Features

Grok 2 is almost completely uncensored and lacks significant guardrails, unlike other models. While we aim to keep this show safe for work, it is worth noting that many online users are experimenting with its capabilities in ways that other models restrict.

Stay tuned for more AI updates.

Grok-2 Is Out and It’s Uncensored!: image 1
Grok-2 Is Out and It’s Uncensored!: image 2
Grok-2 Is Out and It’s Uncensored!: image 3
Grok-2 Is Out and It’s Uncensored!: image 4
Grok-2 Is Out and It’s Uncensored!: image 5
Grok-2 Is Out and It’s Uncensored!: image 6
Grok-2 Is Out and It’s Uncensored!: image 7
Grok-2 Is Out and It’s Uncensored!: image 8
Grok-2 Is Out and It’s Uncensored!: image 9
Grok-2 Is Out and It’s Uncensored!: image 10
Grok-2 Is Out and It’s Uncensored!: image 11
Grok-2 Is Out and It’s Uncensored!: image 12
[ Library ]

You might also enjoy these related resources:

Get more resources
Resources

5 Point Strategy Guide to Prevent MS Copilot Hallucinations and Bad Answers

This guide designed to help teams working on GenAI Initiatives gives you five actionable strategies for RAG pipeline that will improve answer quality and prevent hallucinations.

Resources

IT Survey on 2025 Outlook: The State of Enterprise GenAI and Unstructured Data

How to build ethical, fair, and trustworthy GenAI solutions free of bias and toxicity.

Resources

Rethinking KM in the Age of GenAI

Watch this on-demand webinar produced by KMWorld

Get Demo